Missing carbon reductions? Exploring rebound and backfire effects in UK households
Highlights
► Policy-makers should be mindful of the rebound effect when developing strategies. ► Due to rebound, only around two thirds of expected GHG reductions may be achieved. ► Re-use of avoided expenditure is critical; in extreme case backfire may occur. ► Higher savings reduce rebound: ‘green’ investments minimise rebound. ► Theoretically negative rebound is possible through ‘green’ technology investment.
Introduction
The UK has a target to reduce greenhouse gas (GHG) emissions by at least 80% below 1990 levels by 2050 (HM Government, 2008). It is relying on households to play a pivotal role in meeting this target by encouraging a range of measures including, for example, household energy efficiency improvements.
It is commonly assumed that historical improvements in energy efficiency have reduced energy consumption and associated GHG emissions below the level at which it would have been without those improvements. Nevertheless, before the recession, it was apparent that such improvements failed to reduce energy consumption in absolute terms. Thus while the energy intensity of industrial economies steadily fell, absolute energy consumption attributable to UK households continued to rise, along with the associated carbon emissions (Druckman et al., 2008, Wiedmann et al., 2008, Druckman and Jackson, 2009a).
The most common explanation for the failure to decouple energy consumption and carbon emissions from economic growth is that we have not tried hard enough: energy and carbon prices are too low and policies to encourage energy efficiency and/or lifestyle changes are often small-scale, under-funded, poorly designed and ineffectual. In this view, the appropriate solution is to reinforce these policies—namely, to introduce more regulations, standards, financial support and information programmes alongside the pricing of carbon emissions.
But an additional explanation for the failure to reduce energy consumption is that many of the potential energy savings have been ‘taken back’ by various behavioural responses which are commonly grouped under the heading of rebound effects. While generally neither anticipated nor intended, these effects reduce the size of the energy savings achieved. An example of a rebound effect would be the driver who replaces a car with a fuel-efficient model, only to take advantage of its cheaper running costs to drive further and more often. Some authors argue that these effects lead to increased energy demand over the long term—an outcome that has been termed ‘backfire’ (Saunders, 1992, Brookes, 2000).
Since energy efficiency improvements reduce the effective price of energy services such as travel, the consumption of those services may be expected to increase, thereby offsetting some of the predicted reduction in energy consumption. This so-called direct rebound effect was first studied by Khazzoom (1980) and has since been the focus of much research (Greening et al., 2000, Sorrell and Dimitropoulos, 2007, Sorrell and Dimitropoulos, 2008, Sorrell et al., 2009). But even if there is no direct rebound effect for a particular energy service (e.g. even if consumers choose not to drive any further in their fuel efficient cars), there are a number of other reasons why the economy-wide reduction in energy consumption may be less than simple ‘engineering’ calculations suggest. For example, the money saved on motor-fuel consumption may be spent on other goods and services that also require energy to provide. Depending upon the nature, size and location of the energy efficiency improvement, these so-called indirect rebound effects can take a number of forms (Sorrell, 2007).
The overall or economy-wide rebound effect from an energy efficiency improvement represents the sum of these direct and indirect effects and is normally expressed as a percentage of the expected energy savings. Hence, an economy-wide rebound effect of 20% means that one fifth of the potential energy savings are ‘taken back’ through one or more of the above mechanisms. A rebound effect that exceeds 100% means that the energy efficiency improvements ‘backfire’: in other words, they increase overall energy consumption.
The quantification of rebound effects is difficult, owing to limited data, endogenous variables, uncertain causal relationships, trans-boundary effects and other factors (Sorrell, 2007). As a result, the existing literature is patchy and most studies focus upon only a subset of the relevant effects measured over relatively short time horizons (Sorrell, 2007). While rebound effects are most commonly estimated in relation to energy consumption, they may equally be estimated for carbon dioxide emissions or greenhouse gas (GHG) emissions. The percentage effect may not be the same in each case, owing to variations in the energy, carbon dioxide and GHG intensity of different goods and services. In this paper, we estimate rebound effects in relation to GHG emissions, since we consider the reduction of these emissions to be the primary policy goal.
Most studies of rebound effects focus upon household energy services such as heating and lighting and examine the effect of improving the efficiency of delivering those services—for example, using less electricity to provide the same level of lighting through the replacement of incandescent bulbs with compact fluorescents. However, an entirely analogous effect may occur when individuals choose to change their consumption patterns, with the primary or secondary aim of reducing their environmental impacts or ‘carbon footprint’. For example, individuals may choose to walk or cycle rather than using a car, or to turn off the lights in unoccupied rooms. In these circumstances, the money saved by reduced consumption of the relevant energy service(s) will generally be spent on other goods and services. However, there will be energy consumption and carbon emissions associated with the purchase of these other goods and services. In other words, there will be indirect rebound effects that will offset some (or in extreme cases all) of the intended energy and emissions savings. However, there will not be any direct rebound effects in these circumstances as the household has voluntarily chosen to consume less of that specific energy service.
In this paper, reducing consumption of a particular good or service is termed an abatement action. This is distinct from improving the efficiency of providing a particular energy service which frequently leads to increased consumption of that service and hence a direct rebound effect. So while efficiency improvements lead to both direct and indirect rebound effects, abatement actions lead to only indirect rebound effects. In both cases, these rebound effects are unintended and usually unacknowledged, but their net effect will be to reduce the environmental benefits of the relevant action. Since abatement actions are visible, simple and low cost they are widely promoted by government bodies and non-governmental organisations (NGOs) as an effective means of reducing GHG emissions, as well as being widely practised by individual households. But the indirect rebound effects associated with these actions remain largely unexplored.
This study makes some preliminary estimates of the rebound effects associated with representative abatement actions that may be taken by an average UK household. We consider three actions that have the primary or secondary objective of reducing GHG emissions, namely:
- •
reducing internal temperatures by 1 °C by means of lowering the thermostat;
- •
reducing food expenditure by one third by eliminating food waste; and
- •
walking or cycling instead of using a car for trips of less than 2 miles.
We assume that expenditure avoided due to these actions is either re-spent on other goods and services or is saved. Savings may either be treated as deferred consumption or as a source of funds for investment, but in either case, they will also be associated with GHG emissions. In this paper we treat them as investment funds. We set up a generalised framework in which we can vary the proportion of avoided expenditure that is re-spent or saved, and also vary the expenditure categories in which the re-spending is carried out. The latter may either be in accordance with the estimated expenditure elasticities for the relevant good or service (see below), or determined exogenously in order to estimate upper and lower bounds of the rebound effect.
Four features of this study should be noted. First, unlike other rebound studies, our study takes account of the impact of household savings and investments. This allows us to investigate situations where households put aside rather than re-spend money saved through reduced consumption. Second, we focus specifically on household actions that do not require capital outlay, thereby removing the need to account for the financial and energy consequences of capital investment. Third, we investigate abatement actions involving reduced consumption rather than improved energy efficiency which means that we can focus solely upon ‘income effects’ and ignore any price-induced ‘substitution effects’.1 Finally, we also ignore any ‘general equilibrium’ effects that may result from the abatement actions, such as changes in the price of energy that may induce behavioural changes by other households.
We therefore expect our estimates of the size of rebound effects to be relatively conservative. The rationale for these choices is to produce a simple and transparent study which clearly demonstrates the importance of such effects. Modelling additional dimensions of the rebound effect is the focus of ongoing work.
Our paper is organised as follows: in Section 2 we present the background to estimating the rebound effect, and review relevant studies. Section 3 is the methodology and we present our results in Section 4. Consideration of the limitations of the study are given in Section 5. In the concluding section (Section 6) we summarise our findings and discuss their implications for sustainable consumption policy-making.
Section snippets
Background
Two sets of information are required to estimate the rebound effects from energy efficiency improvements and/or abatement actions by households: first, estimates of the energy consumption and/or GHG emissions that are associated with different categories of household goods and services, and investments; Second, estimates of how the share of expenditure on different goods and services, and the level of savings, varies as a function of prices, income and other variables. The former may be derived
Methodology
The approach taken in this study is straightforward. We first identify three simple actions that an average UK household may take to reduce the emissions attributable to its expenditure, based on suggestions from websites sponsored by the UK government.5 From these we estimate the expected (hoped for) annual reduction in GHG emissions (ΔH), and approximate annual expenditures (Δy) that are likely to be
Household GHG emissions
To set the scene we first examine the estimated expenditure and GHG emissions of an average UK household in 2008. Fig. 1a–c illustrates that whereas, for example, gas accounts for only around 1% of total expenditure, it is the category with the highest GHG emissions. The savings category, in contrast, has a relatively low GHG intensity.
Estimation of rebound under varying conditions
The ‘behaviour as usual’ rebound is estimated by assuming that avoided expenditure is spent and invested in line with current behaviour patterns (as given by Eq.
Limitations of the study
The abatement actions investigated above have been specifically chosen for their simplicity, in that they do not require household capital expenditure and do not lead to any price-induced substitution effects. This makes estimation of the rebound effect simpler and more transparent. Nevertheless, the study has a number of important limitations.
A major limitation of the study is the relatively small number of expenditure categories modelled. These are based on the 12 major COICOP categories
Discussion
Behavioural changes by households are anticipated to make an important contribution to reducing UK GHG emissions. But while policy-makers are increasingly recognising that rebound effects will offset some of the anticipated emission reductions, the empirical evidence on the size of such effects remains very poor. Our study therefore aims to estimate the size of the rebound effect for a set of simple GHG abatement actions advocated by the UK government. These actions have no associated capital
Acknowledgements
We are grateful for the comments and suggestions made by two anonymous reviewers. The research is supported by funding for the ESRC Research Group on Lifestyles Values and Environment (RESOLVE) (Grant number RES-152-25-1004) and also by funding for the Sustainable Lifestyles Research Group (SLRG) by the UK Department for Food and Rural Affairs, ESRC and the Scottish Government.
References (54)
The sufficiency strategy: would rich-world frugality lower environmental impact?
Ecological Economics
(2008)“Green” consumption—no solution for climate change
Energy
(2004)- et al.
Increased energy efficiency and the rebound effect: effects on consumption and emissions
Energy Economics
(2007) Energy efficiency fallacies revisited
Energy Policy
(2000)- et al.
The carbon footprint of UK households 1990–2004: a socio-economically disaggregated, quasi-multiregional input–output model
Ecological Economics
(2009) - et al.
Measuring progress towards carbon reduction in the UK
Ecological Economics
(2008) - et al.
GHG reduction potential of changes in consumption patterns and higher quality levels: evidence from Swiss household consumption survey
Energy Policy
(2009) - et al.
Taxation of multiple greenhouse gases and the effects on income distribution: a case study of the Netherlands
Ecological Economics
(2008) - et al.
Economic, energy and greenhouse emissions impacts of some consumer choice, technology and government outlay options
Energy Economics
(2002) An empirical study on the rebound effect considering capital costs
Energy Economics
(2008)
The inherent dangers of using computable general equilibrium models as a single integrated modelling framework for sustainability impact assessment. A critical note on Bohringer and Loschel (2006)
Ecological Economics
The rebound effect: microeconomic definitions, limitations and extensions
Ecological Economics
Empirical estimates of the direct rebound effect: a review
Energy Policy
Quantifying the global and distributional aspects of American household carbon footprint
Ecological Economics
Indirect and direct energy requirements of city households in Sweden: options for reduction, lessons from modeling
Journal of Industrial Ecology
CGE models and development policy analysis: problems, pitfalls, and challenges
American Journal of Agricultural Economics
Beyond Growth—The Economics of Sustainable Development
Economics and Consumer Behaviour
Cited by (309)
Digital economy to encourage sustainable consumption and reduce carbon emissions
2024, Journal of Cleaner ProductionEarth vitality: An integrated framework for tracking Earth sustainability
2024, Geography and SustainabilityThe Circular Rebound Tool: A tool to move companies towards more sustainable circular business models
2023, Resources, Conservation and Recycling AdvancesEnergy efficiency lags and welfare boons: Understanding the rebound and welfare effects through China's urban households
2023, Renewable and Sustainable Energy ReviewsAchieving emission reductions without furthering social inequality: Lessons from the 2007 economic crisis and the COVID-19 pandemic
2023, Energy Research and Social ScienceDigital economy and carbon rebound effect: Evidence from Chinese cities
2023, Energy Economics